This R03 application, in response to PA-08-214 (Methodology and Measurement in the Behavioral and So- cial Sciences), requests support to develop and validate an implicit assessment tool for estimating alcohol- related self-identity. Implicit assessments of identity related to risky behavior, such as drinking and smoking, predict future risky behavior above and beyond explicit assessments. Unfortunately, there are no easily admin- istered tools for assessing implicit alcohol-related self-identity;consequently, risk assessment and prevention programs in locations such as colleges are unlikely to use available assessments despite possible benefits. In light of ongoing alcohol abuse problems among college students and others, an easily administered risk as- sessment could prove useful for both research and practice purposes. To that end, the proposed research seeks support specifically to: (1) develop the Alcohol Identity Implicit Associations Test (AI-IAT), to measure the strength of the association an individual holds between alcohol and him- or herself;(2) examine the reliabil- ity of the AI-IAT (i.e., test-retest stability;internal consistency);and (3) examine the validity of the AI-IAT (i.e., convergent, divergent, predictive, incremental). The proposed research will complete these tasks using a co- hort of 300 college students, whom we will assess at baseline and at 3-month and 6-month follow-ups. We also will complete a two-week test-retest assessment on 20% of our sample. The establishment of an implicit measure of alcohol-related self-identity, the AI-IAT, will provide the field with several research benefits and lay the foundation for additional identity-related initiatives that hold practical importance in a number of real world settings. It is our intention to follow the proposed research study with an R01 application that extends both the measure-development work described in the current application (i.e., to drugs and gambling) and the settings in which these measures will be tested (i.e., to clinical settings). Automated risk assessment via the AI-IAT will create few demands for prevention specialists and clinicians in practice, while simultaneously yielding vital in- formation for effective treatment and prevention planning. PUBLIC HEALTH RELEVANCE: Alcohol abuse is the greatest single contributor to morbidity and mortality among the college student population. The unchanging prevalence of alcohol abuse among college students in recent years, despite increasing prevention efforts, suggests the need for novel approaches to risk assessment. Most approaches to risk assessment for alcohol abuse among college students require college students to self-report about the antecedents of their drinking behavior. Such assessments present two main limitations: (1) they are relatively insensitive to implicit cognitive structures, which operate outside of conscious awareness;and (2) they are susceptible to students'desires to present themselves in a positive light. The proposed research will develop and validate a new risk assessment tool that is less susceptible to self-presentation desires and more sensitive to an important implicit construct, alcohol-related self-identity. Evidence suggests that alcohol-related self- identity may be an important antecedent to risky drinking behavior. Therefore, the development of this tool will improve the ability to prospectively assess risky drinking behavior among college students, leading to more effective early intervention programs. Additional adaptations could easily extend this tool to clinical populations, improving treatment providers'abilities to estimate for individual clients the likelihood of treatment success.